MCPcopy Create free account

hub / github.com/Gen-Verse/OpenClaw-RL / functions

Functions10,561 in github.com/Gen-Verse/OpenClaw-RL

↓ 1 callersFunctionmain
Fetch the oldest release of a GitHub repository and download its assets.
Megatron-LM/tests/test_utils/python_scripts/download_unit_tests_dataset.py:88
↓ 1 callersFunctionmain
()
Megatron-LM/tests/functional_tests/python_test_utils/test_optimizer_grads_match.py:327
↓ 1 callersFunctionmain
Run and validate inference tests.
Megatron-LM/tests/functional_tests/test_cases/gpt/gpt_dynamic_inference_tp1_pp1_583m_cuda_graphs_validation/cuda_graphs.py:238
↓ 1 callersFunctionmain
()
Megatron-LM/tests/unit_tests/test_api_backwards_compat_setup.py:139
↓ 1 callersFunctionmain
()
Megatron-LM/tests/unit_tests/find_test_cases.py:51
↓ 1 callersFunctionmain
()
Megatron-LM/examples/mimo/avlm_inference.py:129
↓ 1 callersFunctionmain
()
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/perspective_api.py:92
↓ 1 callersFunctionmain
Main program.
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/generate_samples_gpt.py:232
↓ 1 callersFunctionmain
()
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/annotations/filter-selfgeneration.py:37
↓ 1 callersFunctionmain
()
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/annotations/perspective_api_annotate.py:160
↓ 1 callersFunctionmain
Main program.
Megatron-LM/examples/inference/gpt/gpt_static_inference.py:145
↓ 1 callersFunctionmain
( engine: DynamicInferenceEngine, requests: List[Request], port: int, sampling_params: Samplin
Megatron-LM/examples/inference/gpt/gpt_dynamic_inference_with_coordinator.py:42
↓ 1 callersFunctionmain
()
Megatron-LM/examples/inference/gpt/gpt_dynamic_inference.py:408
↓ 1 callersFunctionmain
Main program.
Megatron-LM/examples/inference/t5/simple_t5_batch_inference.py:102
↓ 1 callersFunctionmain
Compare vision model outputs between mcore and HF given the same fixed input.
Megatron-LM/examples/multimodal/model_converter/vision_model_tester.py:89
↓ 1 callersFunctionmain
Run OCRBenchV2 evaluation.
Megatron-LM/examples/multimodal/evaluation/evaluate_ocrbench_v2.py:71
↓ 1 callersFunctionmain
Run MMMU evaluation.
Megatron-LM/examples/multimodal/evaluation/evaluate_mmmu.py:112
↓ 1 callersFunctionmain
Run RD-TableBench evaluation.
Megatron-LM/examples/multimodal/evaluation/evaluate_rd_tablebench.py:67
↓ 1 callersFunctionmain
()
slime/tools/convert_hf_to_int4.py:55
↓ 1 callersFunctionmain
()
slime/tools/convert_k2_thinking_int4_to_bf16.py:185
↓ 1 callersFunctionmain
()
slime/tools/convert_hf_to_torch_dist.py:152
↓ 1 callersFunctionmain
()
slime/tools/merge_lora_adapter.py:21
↓ 1 callersFunctionmain
(args)
slime/tools/convert_to_hf.py:27
↓ 1 callersFunctionmain
(fp8_path, bf16_path)
slime/tools/fp8_cast_bf16.py:42
↓ 1 callersFunctionmain
()
slime/tools/convert_hf_to_int4_direct.py:306
↓ 1 callersFunctionmain
Remark: Can use `lslocks` to debug
slime/tests/ci/gpu_lock_exec.py:12
↓ 1 callersFunctionmain
()
slime/examples/tau-bench/tau1_mock.py:11
↓ 1 callersFunctionmake_attention_mask
Returns a 2-dimensional (2-D) attention mask :param source_block: 1-D array :param target_block: 1-D array
Megatron-LM/megatron/legacy/data/biencoder_dataset_utils.py:17
↓ 1 callersMethodmake_buckets_equal
Makes as many buckets as unique CP sizes needed. This keeps sample IDs tethered to their sequence lengths throughout the bucketing pr
Megatron-LM/megatron/core/pipeline_parallel/hybrid_cp_schedule.py:55
↓ 1 callersFunctionmake_combine_loss_fn
Build a Fireworks-compatible loss function with pre-computed advantages. Args: old_logprobs: Per-token rollout log-probs for each sample
openclaw-fireworks/fireworks_loss.py:23
↓ 1 callersFunctionmake_dataset
Generates a list of samples of a form (path_to_sample, class). Args: directory (str): root dataset directory class_to_idx (Dict[st
Megatron-LM/megatron/legacy/data/image_folder.py:64
↓ 1 callersMethodmake_dataset
( directory: str, class_to_idx: Dict[str, int], data_per_class_fraction: float,
Megatron-LM/megatron/legacy/data/image_folder.py:178
↓ 1 callersFunctionmake_test_packed_padded_seq_params
(sequence_length)
Megatron-LM/tests/unit_tests/transformer/test_attention_packed_seq.py:30
↓ 1 callersFunctionmake_test_packed_seq_params_with_padding
Create PackedSeqParams with both regular and padded cu_seqlens for testing padded sequences.
Megatron-LM/tests/unit_tests/transformer/test_multi_latent_attention.py:57
↓ 1 callersFunctionmamba_builder
(args, pre_process, post_process, vp_stage=None, config=None)
Megatron-LM/mamba_builders.py:11
↓ 1 callersFunctionmamba_layer_flops
Calculate FLOPs for a Mamba layer.
Megatron-LM/megatron/training/training.py:213
↓ 1 callersMethodmamba_state_shapes_per_request
Returns the Mamba conv and ssm states shapes per input sequence if this block contains Mamba layers (this may not be the case with PP
Megatron-LM/megatron/core/ssm/mamba_block.py:191
↓ 1 callersMethodmamba_state_shapes_per_request
Returns the Mamba conv and ssm states shapes per request.
Megatron-LM/megatron/core/ssm/mamba_mixer.py:1011
↓ 1 callersMethodmapping_registry
(self)
slime/slime_plugins/megatron_bridge/qwen3_5.py:85
↓ 1 callersFunctionmasked_max
Masked max over valid tokens (loss_mask == 1). Returns 0 when mask is empty.
slime/examples/train_infer_mismatch_helper/mis.py:32
↓ 1 callersFunctionmasked_min
Masked min over valid tokens (loss_mask == 1). Returns 0 when mask is empty.
slime/examples/train_infer_mismatch_helper/mis.py:22
↓ 1 callersMethodmatch_graph_config
Matches the best CUDA graph batch dimension for the given real batch dimension. Args: real_batch_dim: The real batch dim
Megatron-LM/megatron/core/inference/batch_dimensions_utils.py:351
↓ 1 callersMethodmax_allowed_chunks
Returns the maximum allowed chunks for this ShardedTensor.
Megatron-LM/megatron/core/dist_checkpointing/mapping.py:236
↓ 1 callersFunctionmaybe_copy
(attribute)
Megatron-LM/megatron/core/tensor_parallel/layers.py:126
↓ 1 callersMethodmaybe_create
(args)
slime/slime/utils/metric_utils.py:128
↓ 1 callersMethodmaybe_init_gloo_group
Create Gloo groups.
Megatron-LM/megatron/core/dist_checkpointing/strategies/two_stage.py:148
↓ 1 callersMethodmaybe_inject
Method that decides whether to inject an error.
Megatron-LM/megatron/core/rerun_state_machine.py:1205
↓ 1 callersFunctionmaybe_log_training_metrics
Log training metrics if writers are available. Args: group_stats: RolloutStats object to pass to writers. current_iteration: Curr
Megatron-LM/megatron/rl/rl_utils.py:832
↓ 1 callersMethodmaybe_miscompare
Method that introduces mismatching results during reruns when an error is injected. When no error is injected, this method defers to the user
Megatron-LM/megatron/core/rerun_state_machine.py:1223
↓ 1 callersFunctionmaybe_prepare_qwen35_text_model
(model_path: str, *, language_only: bool)
slime/slime/backends/sglang_utils/qwen3_5.py:22
↓ 1 callersFunctionmaybe_save_dataloader_state
Saves dataloader state if the dataloader supports it. Currently, this is only used by Megatron Energon dataloader (multimodal) to store its state
Megatron-LM/megatron/training/checkpointing.py:764
↓ 1 callersFunctionmaybe_set
(attribute, value)
Megatron-LM/megatron/core/tensor_parallel/layers.py:115
↓ 1 callersFunctionmerge_datasets
(idir)
Megatron-LM/tests/unit_tests/data/test_preprocess_mmdata.py:46
↓ 1 callersFunctionmerge_datasets
(idir)
Megatron-LM/tests/unit_tests/data/test_preprocess_data.py:66
↓ 1 callersFunctionmerge_embedding_blocks
Merge individual embedding blocks into a single binary mmap file. The embeddings are initially stored in block-sized (e.g., ~100k embeddings per
Megatron-LM/megatron/core/datasets/retro/index/build.py:87
↓ 1 callersFunctionmerge_global_slice_with_shape
Intersects the global slice with the actual shape (prevent overflow).
Megatron-LM/megatron/core/dist_checkpointing/strategies/tensorstore.py:80
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_realworldqa.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_ai2d.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_spdocvqa.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_video_phys_game_bench.py:7
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_ocrbench.py:7
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_video_mvbench.py:7
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_chartqa.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_infovqa.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_vqav2.py:58
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_mathvista.py:10
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_textvqa.py:8
↓ 1 callersFunctionmerge_input_files
Merge input files to a format compatible with the evaluator.
Megatron-LM/examples/multimodal/evaluation/evaluate_video_motionbench.py:9
↓ 1 callersMethodmerge_shards_and_save
(self)
Megatron-LM/megatron/legacy/data/realm_index.py:87
↓ 1 callersFunctionmetadata_finalize_fn
()
Megatron-LM/megatron/core/dist_checkpointing/serialization.py:416
↓ 1 callersMethodmetrics
(self)
Megatron-LM/megatron/rl/agent/api.py:108
↓ 1 callersFunctionmlp_layer_flops
Calculate FLOPs for an MLP layer.
Megatron-LM/megatron/training/training.py:184
↓ 1 callersFunctionmmmu_eval
Run MMMU evaluation.
Megatron-LM/examples/multimodal/evaluation/evaluate_mmmu.py:86
↓ 1 callersFunctionmodel_forward
(model: torch.nn.Module, config: TransformerConfig, micro_batch_size: int = 2)
Megatron-LM/tests/unit_tests/post_training/test_modelopt_module_spec.py:31
↓ 1 callersFunctionmodel_provider
Builds the model. Args: pre_process (bool, optional): Set to true if you need to compute embedings. Defaults to True. post_proces
Megatron-LM/tools/retro/text_generation/retro_text_generation.py:28
↓ 1 callersFunctionmodel_provider
Build and return a simple GPT model for demonstration. Returns: GPTModel: A small GPT model with 2 layers for testing.
Megatron-LM/examples/run_simple_mcore_train_loop.py:54
↓ 1 callersFunctionmodel_provider
Build the model.
Megatron-LM/examples/export/trtllm_export/single_device_export/gpt_single_device_cpu_export.py:30
↓ 1 callersFunctionmodel_provider
Build the model.
Megatron-LM/examples/export/trtllm_export/distributed_export/gpt_distributed_gpu_export.py:29
↓ 1 callersMethodmodel_provider
( self, pre_process=True, post_process=True, layer_spec_fn=get_gpt_layer_with_
Megatron-LM/tests/unit_tests/test_fp8_param.py:83
↓ 1 callersFunctionmodel_provider_llava_avlm
Build a LLaVA-style Audio-Vision-Language MIMO model composed of: • Vicuna language model. • Whisper audio encoder. • CLIP ViT-L/14 v
Megatron-LM/examples/mimo/model_providers/llava_avlm.py:31
↓ 1 callersFunctionmodify_underlying_storage
Replace the underlying raw data of a tensor with new data.
Megatron-LM/megatron/core/fp8_utils.py:478
↓ 1 callersFunctionmoe_layer_flops
Calculate FLOPs for an MoE layer.
Megatron-LM/megatron/training/training.py:189
↓ 1 callersFunctionmonkey_patch_torch_dist
()
slime/slime/utils/reloadable_process_group.py:15
↓ 1 callersFunctionmpu_cross_entropy
(batch_size, seq_length, vocab_size, logits_scale, seed)
Megatron-LM/megatron/legacy/mpu/tests/test_cross_entropy.py:31
↓ 1 callersFunctionnew_get_microbatch_id_in_model_chunk
Helper method to get the microbatch_id within model chunk given the iteration number.
Megatron-LM/tests/unit_tests/pipeline_parallel/test_helpers.py:78
↓ 1 callersFunctionnew_is_first_microbatch_for_model_chunk
Check if an iteration is the first for a model chunk.
Megatron-LM/tests/unit_tests/pipeline_parallel/test_helpers.py:84
↓ 1 callersFunctionnew_is_last_microbatch_for_model_chunk
Check if an iteration is the last for a model chunk.
Megatron-LM/tests/unit_tests/pipeline_parallel/test_helpers.py:91
↓ 1 callersMethodnext_hdp_group
Given a list of (sample_id, sequence_length) tuples, this function aims to assign sequences in a group such that all GPUs in the DPxC
Megatron-LM/megatron/core/pipeline_parallel/hybrid_cp_schedule.py:104
↓ 1 callersMethodnext_iter
Increment current training/validation iteration.
Megatron-LM/megatron/core/full_cuda_graph.py:196
↓ 1 callersFunctionnon_loss_data_func
Callback to compute the acceptance length.
Megatron-LM/examples/post_training/modelopt/finetune.py:454
↓ 1 callersFunctionnormal_forward
(input)
Megatron-LM/tests/unit_tests/tensor_parallel/test_random.py:204
↓ 1 callersFunctionnormalize_msg
(msg, options)
gui-rl/desktop_env/evaluators/metrics/thunderbird.py:148
↓ 1 callersFunctionnormalized_levenshtein_distance
(s1: str, s2: str)
Megatron-LM/examples/multimodal/evaluation/evaluate_vqav2.py:29
↓ 1 callersFunctionnum_released_microbatches
Helper method to count number of released (i.e. popped from input_tensors) microbatches for a model chunk.
Megatron-LM/megatron/core/pipeline_parallel/schedules.py:1112
↓ 1 callersFunctionnum_samples_to_block_ranges
Split a range (length num_samples) into sequence of block ranges of size block_size. Args: config (RetroPreprocessingConfig): Retro p
Megatron-LM/megatron/core/datasets/retro/index/utils.py:34
↓ 1 callersMethodnumel_low_level_dataset
Abstract method implementation For GPT, the underlying IndexedDataset should be split by sequence, as opposed to, say, BERT, which sh
Megatron-LM/megatron/core/datasets/gpt_dataset.py:128
↓ 1 callersMethodoccupy_vm
Mark the path of VM occupied by the pid.
gui-rl/desktop_env/providers/base.py:72
↓ 1 callersFunctionocrbench_v2_eval
Run OCRBenchV2 evaluation.
Megatron-LM/examples/multimodal/evaluation/evaluate_ocrbench_v2.py:37
↓ 1 callersMethodoffload
Offload.
Megatron-LM/megatron/core/pipeline_parallel/fine_grained_activation_offload.py:216
↓ 1 callersMethodoffload_to_cpu
Move optimizer state to CPU to free GPU memory during inference.
Megatron-LM/megatron/core/optimizer/optimizer.py:1414
← previousnext →4,601–4,700 of 10,561, ranked by callers